Fruit quality detection and classification using Computer Vision Techniques
T. Thomas Leonid, Hemamalini, T Nanthine, Krithika
Abstract
Human health has been affected by the use of pesticides, fertilizers, and harmful steroids, which may cause serious health issues. We need a reliable method to identify infections and chemicals in the food we eat since harmful chemicals penetrate the internal organs through fruits and vegetables. The current maximum limit for insecticides that can be taken by people and animals is set by the MRL (i. e Maximum Residue Level). Pollutants are considered to be present if the fruit is found to fall outside of the MRL's acceptable limits. By using Internet of Things (IoT), pesticide concentrations can be measured and recorded on a cloud platform. Keras is also used for distributed training of deep learning models, Tensor Flow is the framework used in the Keras library for software designing purposes. An effective monitoring system must be customized to provide reliable network about diseases in fruits and the occurrence of dangerous pesticides in fruits to the cloud platform that processes and sends them to the application on a user's Android smartphone. Hardware and software design is implemented to obtain an accurate result. Accuracy, Precision and perfection are used to create performance measurements with the help of Performance metrics which is established. Using sensing devices like temperature, gas, pH, moisture, Rasperrypy and a Wi-Fi module, a system was constructed to collect data on the existence of pollutants in the environment. The proposed system is a simple, accurate, and real-time best solution and easy to implement.